1. Field of the Invention
The present invention relates to polarizing glasses and in particular to a polarizing glass used as a polarizer in, for example, a compact optical isolator for optical communication etc.; an optical switch composed of liquid crystal, an electro-optic crystal, a Faraday rotator, and so forth; or an electro-magnetic sensor.
2. Description of Related Art
It is known that a glass having anisotropically shaped fine metal particles, such as silver particles or copper particles, oriented and dispersed therein can function as a polarizer because of its ability to change the light absorption wavelength band of the metal particles when the incident polarization direction is changed.
It is also well-known that such a polarizing glass can be produced by reducing a glass containing elongated copper halide particles or a glass containing elongated silver halide particles. For example, Japanese Unexamined Patent Application Publication No. 5-208844 discloses a process of producing a polarizing glass from a glass containing copper halide particles. This process will be described below. First, glass serving as a matrix is made to contain Cu ions and Cl ions. Next, the glass is melted and molded. Then, the glass is cut into an appropriate size, cast into a mold, and heat-treated. By doing so, CuCl microcrystals with grain sizes of 50 to 300 nm are precipitated. Thereafter, the glass material in which the CuCl microcrystals have been precipitated is processed into a slab, which serves as a preform. At a temperature that imparts a glass viscosity in the range of 107 to 1010 Pa·S, the preform and the copper halide particles in the glass are elongated and then heat-treated in a reducing atmosphere. After the copper halide particles are reduced, a polarizing glass containing elongated, anisotropically shaped metallic copper particles is produced.
The polarizing properties of such a polarizing glass are represented in terms of the extinction ratio and the insertion loss defined by the Expressions below.Extinction ratio=−10 Log(□)(P///P⊥)[dB]  Expression (1)Insertion loss=−10 Log(P⊥/Pin)[dB]  Expression (2)
where Pin: intensity of incident light, P⊥: intensity of exit light in the vertical (transmission) direction, and P//: intensity of exit light in the parallel (extinction) direction.
When the extinction ratio is to be measured, a measuring system as shown in FIG. 1 is normally used. Light with a wavelength of 1.31 μm or 1.55 μm emitted from a laser diode (LD) light source is converted via a collimating fiber into collimated light, which is incident upon a Glan-Thompson prism resulting in unidirectional linear polarization and is then incident upon a polarizing glass. Then, the polarizing glass is rotated, and the maximum and minimum amounts of light detected with a power meter are measured to calculate the extinction ratio based on Expression (1).
At this time, when the distance (measurement distance) between the polarizing glass and the power meter is changed, the calculated extinction ratio also changes. In short, the extinction ratio is small for a short measurement distance, whereas the extinction ratio is large for a long measurement distance. This is probably because weakly scattered light radiated from the polarizing glass in all directions is detected by the power meter, decreasing the extinction ratio, in the case of a short measurement distance, whereas less of such scattered light is detected by the power meter, increasing the extinction ratio, in the case of a long measurement distance.
This relationship between the measurement distance and the extinction ratio will be described by providing specific values.
A decrease of X (dB) in the extinction ratio of the polarizing glass, as a result of a decrease from A mm to B mm in the distance (measurement distance) between the polarizing glass and the power meter photodetector having an aperture portion, can be explained based on the schematic diagram shown in FIG. 4. In the case of the long measurement distance A, the amount of scattered light (in terms of photo-detection area) that can be detected by the power meter photodetector in the case of the short measurement distance B decreases in proportion to (B/A)*2. In contrast, when the measurement distance decreases from A to B, the amount of scattered light that can be detected increases in proportion to (A/B)*2.
In summary, assuming that the scattered light intensity is not angle-dependent, the decrease of X (dB) in the extinction ratio of the polarizing glass, as a result of a decrease from A mm to B mm in the distance (measurement distance) between the polarizing glass and the power meter photodetector with an aperture portion, can be roughly calculated from the following Expression as a model, based on the ratio between the power meter photodetector areas that can detect scattered light.X=10 Log(A/B)*2  Expression (3)
Although the actual decrease in the extinction ratio of the polarizing glass slightly deviates from the value calculated from Expression (3) because scattered light is angle-dependent in an actual copper-based polarizing glass, the extinction ratio at a measurement distance of 15 mm decreases to approximately 40 dB, compared with an extinction ratio of 55 dB at a measurement distance of 300 mm.
For these extinction ratio measurements of the polarizing glass, collimated light is used as the source light, with the shortest distance that can be measured by the measuring instrument being limited to a measurement distance of 15 mm, as shown in FIG. 1. In an optical system for use in an actual optical isolator, diverging light from a laser diode (LD) light source is focused by a lens onto the optical isolator. For this reason, the proportion of re-radiated light due to resonance scattering from the polarizing glass surface being accepted by an optical fiber decreases, compared with the optical system featuring collimated light shown in FIG. 1. Therefore, the decrease in the extinction ratio as a result of the measurement distance being decreased is slight. In other words, the values of the extinction ratio of the polarizing glass and the isolation of an optical isolator incorporating this polarizing glass do not always match, though the relationship between the extinction ratio and the isolation is such that as the extinction ratio increases, the isolation also increases.
An investigation was conducted to confirm the above-described fact.
For a copper-based polarizing glass with an extinction ratio of 40 dB at a measurement distance of 15 mm and at a wavelength of 1.55 μm, the extinction ratio measured at a measurement distance of 300 mm and at a wavelength of 1.55 μm was 56 dB. When a free space optical isolator for a wavelength of 1.55 μm was assembled in the same manner as described above to measure its isolation, the result was 36 dB, which demonstrates that the polarizing glass is suitable for use in the free space optical isolator.
On the other hand, when a pigtail optical isolator for a wavelength of 1.55 μm was assembled using the same sample, i.e., the copper-based polarizing glass with an extinction ratio of 40 dB at a measurement distance of 15 mm and at a wavelength of 1.55 μm, together with a commercially available garnet film, permanent magnet, and single mode fiber to measure the isolation of the isolator, the result was 28 dB.
Considering that the isolation of typical optical isolators is specified as 30 dB or more, the isolation of 28 dB was below the specification. That is, the polarizing glass with an extinction ratio of 40 dB or less at a measurement distance of 15 mm cannot be used in a pigtail optical isolator.
Nowadays, pigtail optical isolators in which the polarizing glass is bonded directly to the fiber are widely used as optical isolators for Metropolitan Area Networks. In such optical isolators, the near-field extinction ratio, defined as the extinction ratio for a short distance between the polarizing glass and the power meter, is important.
On the other hand, because it is presumed from the principle of Mie scattering that as the minor-axis diameter of anisotropic fine metal particles in the polarizing glass increases, this scattered light also increases, it is important to produce a polarizing glass containing anisotropic fine metal particles with minimized minor-axis diameters.
From the conclusion described above, to achieve a high near-field extinction ratio value, i.e., a high extinction ratio value for a short distance between the polarizing glass and the power meter, it is advantageous to increase the aspect ratio (major-axis diameter/minor-axis diameter) of the post-reduced metal particles by increasing the degree of stretching of the metal halide particles.
However, there has been a problem in that when metal halide particles are to be stretched by heat-stretching a preform, the preform easily breaks, decreasing the yield. This problem will be described in detail.
If the glass of the preform is hard, i.e., if the glass has a low flexibility, then it is difficult to stretch the preform. As a result, the preform needs to be stretched with a high tensile force to increase the degree of stretching of the metal halide fine particles. It is true that the rate of stretching of metal halide fine particles increases with a high tensile force, but the probability of the preform being broken due to that high tensile force also increases.
On the other hand, if the glass of the preform is soft, i.e., if the glass has a high flexibility, the preform can be stretched with a low tensile force. However, because such a glass has a low mechanical strength and is fragile, the glass may break even with a low tensile force.
Thus, because the required tensile force cannot be applied to the metal halide fine particles, whether the preform glass is hard or soft, in an attempt to stretch the preform with a tensile force such as not to cause glass breakage, it is difficult to increase the degree of stretching of the metal halide fine particles. Therefore, it is not possible to increase the near-field extinction ratio, as described below in detail.
To achieve a high extinction ratio, a tensile force that is almost as high as that at which the preform glass breaks is normally applied to stretch the preform in the stretching process. However, the rate of stretching of CuCl will not surpass a certain level, whether the preform glass is hard or soft, under conditions in which frequent breakage must be avoided in the production process. Consequently, the rate of stretching of the metal halide particles has not been enhanced due to an increase in the probability of fracture.
As described so far, when a known copper-based polarizing glass is to be stretched with a high tensile force in an attempt to increase the rate of stretching, i.e., the aspect ratio (major-axis diameter/minor-axis diameter), of metal halide fine particles in the form of precursors of anisotropic fine metal particles, there is a possibility of the glass being broken during the stretching process. For this reason, the minor-axis diameters of the anisotropic metal particles cannot be decreased if a high yield is to be achieved in the production process. Therefore, the extinction ratio at a measurement distance of 15 mm is as low as about 40 dB. Consequently, there has been a problem in that the isolation of a pigtail optical isolator incorporating a conventional polarizing glass is too low to satisfy an isolation performance of 30 dB or more, which is the specification of typical isolators, even though the same polarizing glass can be satisfactorily used in a free space optical isolator.